# Voxi Lite server image (CPU). Build from the release folder: # docker build -t voxi-lite:1.0.0 . # docker run --rm -p 8080:8080 -e VOXI_THREADS=4 voxi-lite:1.0.0 # One model per container; scale with more containers (about 1.5 GB RAM each). Nothing is downloaded at runtime. FROM python:3.11-slim ENV PYTHONUNBUFFERED=1 PIP_NO_CACHE_DIR=1 \ HF_HUB_OFFLINE=1 TRANSFORMERS_OFFLINE=1 \ VOXI_THREADS=4 OMP_NUM_THREADS=4 \ VOXI_WARM=/app/warm_phrases.json WORKDIR /app COPY requirements.txt . RUN pip install --extra-index-url https://download.pytorch.org/whl/cpu -r requirements.txt COPY . . RUN useradd --create-home voxi && chown -R voxi /app USER voxi EXPOSE 8080 HEALTHCHECK --interval=30s --timeout=5s --start-period=90s \ CMD python -c "import urllib.request,sys; sys.exit(0 if urllib.request.urlopen('http://127.0.0.1:8080/health', timeout=4).status == 200 else 1)" CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "8080", "--no-access-log"]